Browse Skills

A skill for designing game quests using patterns like quest archetypes, reward psychology, difficulty curves, and player motivation frameworks. Useful for quest design, reward systems, and mission structure.

Provides a systematic approach to character design for AI image generation, ensuring visual consistency through character sheets, silhouette testing, expression charts, and prompt engineering strategies.

A full production pipeline where an AI agent team collaborates to create comics from storyboarding to dialogue, image generation, and editing. Supports 4-panel strips, short-form, and long-form formats.

A methodology for comic panel composition, providing layout patterns, camera angles, gaze flow design, and page rhythm. Used by storyboarder and image-generator agents when designing panels and generating images.

Generate photoreal or designed images using OpenAI's gpt-image models through fal.ai. Supports text-to-image and edit-from-reference with gpt-image-1 and gpt-image-2, ideal for character anchors, scene keyframes, and storyboards.

Generates videos from images or text using FAL AI models like Kling, Seedance, or Veo. Routes through the GooseWorks proxy for billing and is ideal for creating base clips for video ads.

Guides packaging of AAMAS multiagent research artifacts—code, environments, opponent sets, seeds, and logs—as anonymous supplements or public releases. Enables game-theory and MARL reviewers to inspect and re-run interaction claims.

Helps AI agents configure the free OpenAI-compatible FreeTheAi API, providing setup for SDKs, chat clients, and tools with guidance on keys, model aliases, and common errors.

Implements input and output validation guardrails for LLM applications to prevent prompt injection, enforce content policies, detect PII, and protect against jailbreaks. Use when deploying AI chatbots or RAG pipelines that need safety controls.

Uses LLMs to generate fuzzing seeds, evolve grammars with coverage feedback, and scale proof-of-vulnerability generation. Ideal for bootstrapping fuzzers like AFL++ and triaging HTTP traffic during authorized pentests.

Test MCP servers and the AI clients that consume them for tool poisoning, prompt injection, credential theft, and supply-chain attacks. Use during authorized security reviews of AI agent integrations to identify vulnerabilities before deployment.

A comprehensive guide for creating effective skills that extend Claude Code's capabilities with specialized knowledge, workflows, and tool integrations. Use this when building new skills or updating existing ones.